dograh
Open Source Voice Agent Platform
Discover how Dograh's realtime turn context manages speech-to-speech sessions using a hybrid approach with global TurnContextManager and Pipecat's turn_var for precise logging and analytics in AI workflows.
How Dograh Handles Transport Setup for Different Telephony ProvidersDograh simplifies transport setup for telephony providers using its provider-registry pattern. Learn how Dograh builds FastAPIWebsocketTransport for seamless integration.
How Dograh AudioConfig Determines Sample Rates and Buffer Sizes for the PipelineDiscover how Dograh AudioConfig sets sample rates and buffer sizes for your voice AI pipeline, ensuring VAD compatibility and efficient 16-bit PCM processing.
How the Disposition Mapper Applies Call Outcome Codes in DograhLearn how the disposition mapper applies call outcome codes in Dograh by translating raw identifiers into customized codes using the DISPOSITION_CODE_MAPPING configuration.
How Knowledge Base Retrieval Integrates with Voice Pipeline Context in DograhDiscover how Dograh’s Pipecat voice pipeline seamlessly integrates knowledge base retrieval with voice pipeline context, binding vector search results to conversational state for enhanced interactions.
How Custom Tools Are Registered and Executed in Dograh’s Voice PipelineLearn how Dograh's voice pipeline registers and executes custom tools. Discover dynamic tool integration with CustomToolManager for LLM-invoked actions.
How the Pipeline Builder Constructs the Complete Voice Processing Flow in DograhDiscover how the Dograh pipeline builder crafts a complete voice processing flow, from audio input and STT to LLM inference, TTS, and output asynchronously.
How Gemini Live Reconnection Handles Tool Results Across Sessions in DograhLearn how Dograh's Gemini Live reconnection handles tool results across sessions. Discover the four-phase state machine preventing data loss and ensuring seamless bot interactions.
How Workflow Layout Reconciliation Works After LLM-Authored Edits in DograhLearn how Dograh reconciles workflow layouts after LLM edits. Discover the three-phase algorithm used to restore node positions before saving changes.
How TURN Credentials Are Configured and Rotated for WebRTC Calls in DograhLearn how Dograh configures and rotates TURN credentials for WebRTC calls using time-limited REST API and HMAC-SHA1 for secure, automatic credential rotation without server state.
How the Dograh Campaign Dispatcher Manages Outbound Calling at ScaleDiscover how the Dograh campaign dispatcher handles large scale outbound calling using ARQ workers, Redis rate limiting, circuit breakers, and Pipecat. Learn to protect APIs and scale effectively.
Have a question about this repo?
These articles cover the highlights, but your codebase questions are specific. Give your agent direct access to the source. Share this with your agent to get started:
curl -s "https://instagit.com/install.md" 
Maintain an open-source project? Get it listed too →